JP0655029

Divertor Experiments with SMBI and Strong Gas Puffing on HL-2A X.R. Duan, X.T. Ding, Q.W. Yang, L.W. Yan, L.H. Yao, W.Y. Hong, W.M. Xuan, D.Q. Liu, L.Y. Chen, X.M. Song, J.H. Zhang, Z. Cao, Z.Y. Cui, W. Li, Yi Liu, YD. Pan , L. Pan, Y. J. Zeng, Y. Zhou, W.C. Mao, Yong Liu and HL-2A team Southwestern Institute of Physics, P.O.Box 432, Chengdu 610041, China Email: [email protected] Abstract In the HL-2A's 2004 experiment campaign pulsed supersonic molecular beam injection (SMBI) and strong hydrogen gas puffing under the configuration were used for gas fueling. The experimental results show that the SMBI of hydrogen can reduce the heat flux to the divertor target plate. The electron temperature measured by the Langmuir probe array decreases during the injection of the molecular beam, whereas the electron density increases significantly. It indicates that the pressure along the magnetic line tends to be constant at a new equilibrium level. Under our experimental conditions the plasma emission at the edge increases during the injection of the hydrogen, but in the divertor region it behaves inversely, the Ha and CIII intensities tend to decrease during the SMBI pulse as the electron temperature measured at the target plates does. This shows that the plasma emission in the divertor under our discharge conditions is more sensitive to the electron temperature than to the electron density. In the divertor plasmas with strong hydrogen gas puffing a high plasma density up to 4.4 x 1019m*3 was achieved. Besides we observed a phenomenon similar to the partially detached divertor regime, which is being studied in the open divertor like DIII-D to reduce the peak heat flux on the target plates near the separatrix. After a strong gas puffing the electron temperature measured on the outer divertor target plate near the separatrix decreases till below 5 eV or even lower, but that of the farther outer divertor target plate does not change obviously; the CIII and Ha emission at the plasma edge decreases as expected, but the Ha emission near the X-point increases. This result reveals some interesting characteristics, which need to be studied by modeling and further experiments.

I Introduction In magnetically confined fusion plasma the plasma-wall interaction has a deleterious effect on plasma performance. Hence, much effort has been devoted to improving wall conditions and modifying the magnetic topology to reduce the influence of the wall. As early as 1951, divertor was proposed by Spitzer to isolate the bulk plasma from the vessel wall[1'. This concept has been realized successfully in ASDEX, where the H-mode (high confinement mode) [2] during high-power neutral beam injection was achieved, and led to a new generation of diverted tokamak experiments. During recent years the success of diverted tokamaks, such as JET,

141 JT-60U, ASDEX Upgrade, and DIII-D, has been remarkable. HL-2A is a divertor tokamak reconstructed at the SWIP in Chengdu based on the original ASDEX main components (vacuum vessel and magnet coils) [3>4] and the HL-2A project is an important part of China's fusion research program. In the last two years' campaign of the HL-2A experiment, divertor plasma was achieved. During the 2004 experiments, the supersonic molecular beam injection(SMBI) and strong hydrogen gas puffing were used for gas fuelling, and some unique phenomena were observed. These results will be discussed in the following sections. This paper is organized as follows. The experimental aspects of the HL-2A tokamak are described in Section II. Section III gives some key experimental results of the divertor plasmas with molecular beam injection and strong gas puffing, and related discussions are also given in this section. At the end there is a short summary. II Experimental Procedures III HL-2A tokamak

The HL-2A tokamak is characterized by a large closed divertor chamber (see Fig.l), and can be operated in double null, upper single null and lower single null configurations with the same main plasma condition. The vacuum vessel, 16 toroidal field coils, poloidal field coil systems and supporting structure of the former ASDEX are adapted for HL-2A141. The other sub-systems of HL-2A, including the pumping system, cooling system, power supply system, diagnostics system, etc. have been or are being constructed. The main pumping system of HL-2A is composed of eight turbo molecular pumps (35001/s each) and two sets of cryopumps with two pre-stage pumps. The divertor pumping system is composed of 18 titanium getter pumps installed in the divertor chambers. The vacuum vessel can be baked up to 130°C-150°C for degassing and a glow discharge device is installed in the vessel for cleaning the inner surface of the vessel. Eight DC pulse power supply units have been constructed for the coil system of the toroidal field (TF), the Ohmic heating (OH), the vertical field(VF), the radial field(RF), the multipole field(MP), the multipole compensation field(MPC), and so on. The divertor of the HL-2A tokamak consists of three multi-pole coils MP1, MP2, MP3 and two neutralized targets, each MP2 coil has eight turns and each MP1/MP3 coil four turns respectively. These coils are used to form the divertor configurations with different methods of power supply. According to the results of the equilibrium analysis by the SWEQU code[5], which is the equilibrium analysis code by solving the Grad-Shavranov equation, the ratio of IMP2/Ip is the most crucial for the divertor operation. According to the simulation results by scanning this ratio from 7% to 10% using the SWEQU code, it has been found that the optimum value of IMP2/Ip is ~8%. The details of the HL-2A tokamak can be found in reference 4. 11.2 Diagnostics A cross section of a diverted HL-2A plasma including the flux surface contour is shown in Fig.l. Some diagnostics are also shown in the figure. To investigate the

142 plasma features in the divertor, five kinds of diagnostics have been mounted in the lower divertor. The microwave interferometer, target plate Langmuir probe arrays and visible spectrometer are used to measure the profiles of the electron density, electron temperature, and Ha emission, respectively. Four target plate Langmuir probe arrays are fixed on the four target plates, respectively. Each array consists of seven probes with three tips, and the vertical distance between two probes is 1.0 cm. The neutral gas pressure is given by fast ionization-gauge at the divertor chamber. In particular, we use the signals detected by 18 pick-up coils located around the plasma column and Current filament (CF) code, a plasma boundary identification code, to construct the plasma LCFS(Last Closed Flux surface). Besides, about 30 diagnostics have been installed in the main chamber of the device, which include the HCN interferometer, ECE, Thomson scattering, CX neutral particle analyzer, bolometer array(16 channels), VUV spectroscopy, reciprocating probes, and the visible spectrometers at the mid-plane of the device. In the divertor experiments on HL-2A, several methods are used to identify the formation of the divertor configuration. A CCD camera is the most direct tool, which can take the images of the cross-section of the plasma discharges. Ill Results and Discussions In the 2004 experiment campaign of the HL-2A tokamak, the SN divertor configuration was in operation. The plasma parameters achieved were: Ip = 320 kA, ne = 4.4 x 1019m"3, Bt = 2.2 T, and plasma duration T = 1580 ms. III.l Divertor plasma with SMBI To achieve a high plasma density and better plasma confinement, high pressure (0.3-0.4Mpa) SMBI was adapted for gas fuelling161. Fig.2 shows a typical picture of the plasma discharge during a SMBI pulse. The discharge conditions were as follows: the plasma current Ip = 180 kA, line averaged electron density ne = 2xl019m"3, and toroidal Bt = 2.1T. Ten SMB pulses were injected into the vacuum chamber via a Laval type nozzle located at the mid-plane of the tokamak's low field side. Each pulse length was 10 ms, and the time interval between two pulses was 20 ms. The first pulse was injected at t = 200 ms. Two bright legs were observed in the lower divertor throats (see Fig.2), which indicated that the plasma had gone into the lower divertor along the magnetic lines. On the left side of the picture there is a bent bright belt, it is due to the SMB injection. Fig.3 provides the experimental results. Ten peaks can be seen clearly from the Ha emission from the plasma edge. This is due to the density increase and temperature decrease at the edge of the plasma during the SMB injection, but the CIII and bremsstrahlung emission decreases obviously after SMB injection because of the decrease in the relative carbon concentration and divertor configuration. After the first pulse the intensity of Ha emission at the mid-plane was lower than its value before the SMB pulse. It was the result of the divertor configuration. During the pulsed SMB injection, the divertor plasma showed some unique results: the electron temperature near the strike point decreased obviously during the SMBI pulses, it indicated a

143 cooling effect of the SMBI under the divertor configuration, i.e. it could reduce the heat load of the divertor target plate; whereas the electron density measured at the same location increased significantly during the pulse. This contrary tendency of the ne and Te variation indicated that the plasma pressure along the magnetic line tended to be constant at a new equilibrium level, the visible emission including Ha and CIII also changed with the pulse, and during the pulse it reached a minimum. This indicated the variation of visible emission in the divertor was determined mainly by the electron temperature under our discharge conditions. During the SMBI the particle confinement time was increased by a factor of 2. III.2 Divertor plasma with strong hydrogen gas puffing Conventional gas puffing was also used for gas fuelling in our experiments. In the 2004 experiment campaign of the HL-2A, gas pulses with different duration and pressure were applied to increase the plasma density, and a high plasma density up to 4.4x1019m'3 was achieved. In our experiments, an interesting phenomenon was observed during the divertor operation. According to the measurement of the CCD camera, which was arranged tangentially to the tokamak torus, two bright legs of the divertor were observed at the beginning of the divertor regime, and the electron temperature measured at different location on the target plates varied in the same manner, and well above 10 eV, i.e., it worked at the attachment regime. After about 50 ms, only the light belt near the outer divertor throat appeared. The plasma density and current remained unchanged, as shown in Fig. 4. Farther out from the separatrix strike point on the outer target plate (see Fig. 5), the electron temperature changed slightly with the discharge time and the electron density changed almost inversely (see Fig. 4(e)), hence the plasma pressure almost kept constant at the location. The Ha and CIII emission from the bulk plasma and bremsstrahlung emission did not vary obviously in the main plasma, but the electron temperature near the separatrix strike point decreased with the time till below 5 eV, at the same time, and the Ha emission measured near the X-point increased. This indicates that the radiation in the vicinity of the X-point is enhanced. From Fig 4 (i) and (j) we can see that the Ha emission and CIII in the divertor decrease as the Te does at z=-82cm. This phenomenon is similar to the partially detached divertor regime observed in the open divertor tokamak such as DIII-D[7-9] In our experiment, only the Ha and CIII in the vicinity of the separatrix strike point was measured; hence the Ha emission had a similar temporal evolution as the Te near the separatrix strike point. According to the results observed, near the inner divertor throat, there was no obvious emission, but the Ha emission in the vicinity of the separatrix was enhanced, i.e., the majority of the radiation was along the outer divertor leg, and the temperature near the outer strike point (OSP) was significantly decreased, the heat flux reduction was largest near the OSP. Furthermore, the electron pressure near the OSP decreased significantly but only modest changes were observed farther out in the scrape-off layer, i.e., these divertor plasmas were detached near the separatrix, but remained attached farther out in the SOL. Most of the characteristics observed in the discharge were in accordance with those observed on DIII-D during

144 the PDD operation. According to the report on DIII-D[9], the PDD regime is characterized by reduced target plate heat flux and ion current near the strike point, enhanced upstream impurity radiation, and low plasma temperature in much of the divertor. In addition, the inner leg could be completely detached and cooled up to the X-point to the temperatures below which neither carbon nor radiate substantially (Te<~leV). but in our case, the plasma density and temperature were relatively low, although the Ha emission near the separatrix increased, no obvious CIII increase was observed. As mentioned above, the HL-2A is characterized by a large closed divertor chamber, there is some difference from the point view of the divertor structure, e.g. the open divertor tokamak has a large private flux region. According to Stangeby[10], the neutrals play a very important role in triggering detachment, especially the neutrals from the private flux region, which enter the outer leg SOL by migrating across the separatrix. Hence on HL-2A no significant radiation enhancement was observed. Here we have only presented some preliminary results concerning the phenomena similar to the PDD , it will be investigated experimentally in more detail by improving some diagnostics near the x-point and in the divertor. IV Summary The divertor experiments with SMBI and strong hydrogen gas puffing have been conducted. The results show that SMBI as a gas fueling technique can improve the particle confinement, and it influences ne and Te in the divertor, and especially can reduce the heat load of the divertor target plate. In the divertor plasmas with strong hydrogen gas puffing a high plasma density up to 4.4 x 1019m"3 has been achieved. Besides, we have observed a phenomenon similar to the partially detached divertor regime, which is being studied in the open divertor tokamaks like DIII-D to reduce the peak heat flux on the target plates near the separatrix. After a strong gas puffing the electron temperature measured on the outer divertor target plate near the separatrix decreases till below 5 eV or even lower, but that of the farther outer divertor target plate does not change obviously; the CIII and Ha emission at the plasma edge decreases as expected, but the Ha emission near the X-point increases. This result has some interesting characteristics, which need to be further studied by modeling and experiments.

References 1 C S Pitcher and P C Stangeby, PPCF 39(1997)779-930 2 The ASDEX Team 1989 Nucl. Fusion 29 1959. 3 J.C. Yan, et al. "Status and Plan of the HL-2A Project", Proceedings of the 19th IEEE/NPSS Symposium on Fusion Engineering, Atlantic City, USA, Jan. 2002. 4 Y.Liu, X.T.Ding, et al. Recent Advances in the HL-2A Tokamak Experiments, 20th IAEA Fusion Energy Conference, Vilamoura, Portugal, 1-6 November 2004.) 5 Li F Z, Xu W B and Shi BR 1989 chin. J. Plasma Phys. 9 12) 6 Lianghua Yao, et al. Nucl. Fusion 38(4) 1998 631

145 7 Petrie T W et al. 1992 J.Nucl. Mater. 196-198, 848 8 FENSTERMACHER M E et al. Proc. Of 12th Conf. Plasma Surface Interactions, St. Raphael, France 1996. 9 W P West et al. PPCF 39(1997) A295-310. 10 Stangeby P C, 1993 Nucl. Fusion 33 (1695)

146 pickup coils

separatnx

core region

- inner di veitor leg -multiple Langmir probes

Fig. 1 A cross section of the HL-2A vacuum vessel with separatrix contour plot of a typical lower single-null divertor plasma configuration is shown along with some important diagnostics. The inner and outer divertor leg locations are also shown in the figure.

FIg.2 Divertor discharge snapshot during the SMBI. The left belt is due to the SMBI pulse into the plasma, the bottom light belts are the plasma emission in the vicinity of the inner and outer .

147 0 200 400 600 800 1000 t (ms) shot 2987#

Fig. 3 Time evolution of a typical SN discharge with pulsed-SMBI beginning at 200 ms. The traces shown are, (a) the plasma current (Ip), (b) electron density(ne), (c) gas pressure, (d) CIII emission in the main plasma, (e) electron density and temperature near the separatrix at z=-83 cm, (f) electron Te and ne at z=-83 cm, (g) Ha emission integrated over the mid-plane line, and (h) Ha emission at the plasma edge.

148 200 400 600 800 1000 t(ms)

Fig. 4 The traces shown are, (a) the plasma current (Ip), (b) electron density(ne), (c) gas pressure, (d) CIII emission in the main plasma, (e) electron density and temperature near the separatrix at z=-83 cm, (f) electron Te and ne at z=-83 cm, (g) Ha emission integrated over the mid-plan line, and (h) Ha emission at the plasma edge.(i)

149 X-prirt Haenissiaru/ slight belt

strike point

iiearOSP father from OSP

Largmirprobes div Fig. 5 Schematic diagram of the lower divertor chamber

150